ElkArte Community Modern, Free, Powerful
Community building forum software

My friend runs a MyBB forum. It is somehow having problems like getting errors when the thread has some 3,000 something replies. Not to that extend but say a Windows 8 thread would be getting a lot of replies. Can ElkArte handle that kind of stuff. I believe it can but just wanted to ask officially.

There's no Elkarte forums that I know of that are big and busy. Since Elkarte is pretty close to SMF, it's safe to assume it can though. So the answer is yes.

Can't we use some sort of spam bot software and test what happens when there is a large no. of posts, threads ?

That's more an issue with pagination which your database is mostly responsible for. To be honest, you could probably overcome that in MyBB with some simple database changes. If you're using MySQL, make sure it is upgraded and query cache is enabled. If it isn't already, make sure that memcached or similar is installed and enabled.

You can definitely attempt to stress test any forum, but most issues aren't from too many topics, messages, members, etc. They are issues of updating rows. In other words, simultaneous posting and accessing. In order to test that programmatically, you need to use something like Apache Bench (ab).

I don't know anything about MyBB, but I know a lot about making forums more performant, especially SMF. Like I said, since Elkarte is largely based on SMF, it is safe to say that they'll run pretty close to the same. There are a ton of very large forums running SMF. Some topics have thousands of posts and run fine.

Yep, pretty much what groundup said.
On my forum I have just a couple of topics with 4000+ messages (the top ten goes from 700 to 4000+), but I'm using a shared hosting with an overseller (or very close to) and I have never noticed any problem.

I'll migrate soon to ElkArte...

I have full faith in ElkArte and the team. I was just implying that if we do such a test before or just after first stable release, it will look more professional.

My largest topic is 58,236 replies (it's on my local DEV board).. no problems.

Not a matter of faith, as explained by groundup, it's difficult to do such benchmarks, and the "live test" is the only reliable test...
Fortunately, the code is not "new", so we have data from the past we can rely on.